The CBD Epilepsy Discovery That Was Ignored for 40 Years
Chronic administration of cannabidiol to healthy volunteers and epileptic patients
In 1980, a Brazilian-Israeli team published the first clinical trial of CBD for epilepsy: 10% of CBD patients still seizing versus 90% of placebo. The finding was ignored for four decades until desperate parents and modern trials vindicated it.
In 2017, when the New England Journal of Medicine published Devinsky's landmark trial showing CBD reduced seizures in Dravet syndrome, it was hailed as a breakthrough — the first rigorous evidence that cannabidiol could treat epilepsy.
It wasn't. The first rigorous evidence had been published thirty-seven years earlier, in a Brazilian pharmacology journal, by a team that included both the pioneer of Brazilian psychopharmacology and the man who had isolated THC. The scientific community ignored it for four decades.
This is the story of how the most important finding in cannabis medicine was discovered, forgotten, rediscovered, and vindicated.
The Forgotten Trial
The paper had nine authors. The senior investigators were Elisaldo Carlini — a Brazilian pharmacologist who had spent three years at Yale before founding the Department of Psychopharmacology at the São Paulo School of Medicine (UNIFESP) — and Raphael Mechoulam at the Hebrew University of Jerusalem, who had isolated THC sixteen years earlier.
Their collaboration had begun in the 1970s, when Carlini noticed something odd about Brazilian cannabis samples. The THC content alone didn't explain all the biological effects he was observing. Some extracts were sedating where pure THC was stimulating. Some were anticonvulsant where THC was not. Other cannabinoids — particularly cannabidiol — were clearly doing something.
In 1973, Carlini's group published preclinical evidence that CBD protected mice against seizure-inducing agents and blocked audiogenic seizures in rats. It also reduced seizure susceptibility in hippocampal tissue. The animal data was clear: CBD was anticonvulsant.
The logical next step was a human trial. It happened in São Paulo. And it worked.
The Study
10% vs. 90%
of patients still experiencing generalized epileptic seizures at the end of treatment — CBD group versus placebo group. In a small trial of 15 drug-resistant epilepsy patients, CBD virtually eliminated generalized seizures in most patients who received it. The placebo group showed almost no improvement.
For comparison, the 2017 NEJM Dravet trial showed 38.9% seizure reduction with CBD vs 13.3% with placebo. The 1980 trial's results were, if anything, more dramatic — though the sample size was far smaller.
Cunha et al. (1980), Pharmacology 21:175-185
The results were remarkable. In a double-blind, placebo-controlled trial — the gold standard even in 1980 — CBD had virtually eliminated generalized seizures in most patients. The safety profile was clean. The methodology was sound.
And then nothing happened.
Forty Years of Silence
The result was a four-decade gap. From 1980 to the 2010s, no major clinical trial of CBD for epilepsy was conducted. The Cunha study sat in the literature, occasionally cited in reviews, never replicated, never followed up.
Rediscovery
The rediscovery didn't come from scientists. It came from parents.
In 2013, the Figi family's story broke on CNN — Charlotte, a girl with Dravet syndrome, dramatically improving on CBD-rich cannabis oil. A 2013 Stanford retrospective study of 19 children with refractory epilepsy found over 80% seizure reduction in 52% of CBD users. The anecdotes accumulated. The desperation was undeniable. And researchers finally went back to the literature and found that a Brazilian team had already demonstrated this — in a double-blind trial — thirty-three years earlier.
Devinsky's expanded access program launched in 2014. The NEJM Dravet trial was published in 2017. The LGS trial followed in 2018. Epidiolex was approved in June 2018.
Mechoulam lived to see the vindication. Carlini, who died in September 2020 at age 90, saw his forgotten work become the historical foundation of an FDA-approved drug.
What This Study Teaches
Myth vs. Reality
CBD for epilepsy was discovered by parents in Colorado in 2013.
Parents in Colorado brought CBD for epilepsy to public attention. Brazilian scientists discovered it in a clinical trial in 1980 — and in animal studies in 1973. The Cunha trial was double-blind, placebo-controlled, and showed dramatic results. It was ignored for forty years due to the War on Drugs, regulatory barriers, the small sample size, and the prevailing assumption that CBD was pharmacologically inert. The modern Epidiolex trials confirmed what Carlini and Mechoulam had demonstrated decades earlier.
The Evidence
Cunha et al. (1980): 10% of CBD patients still seizing vs 90% placebo over 4.5 months. Carlini et al. (1973): CBD anticonvulsant in mice and rat hippocampal tissue. Devinsky et al. (2017): 38.9% seizure reduction in Dravet (NEJM). The timeline shows discovery preceded rediscovery by 37 years.
Cunha et al. (1980), Pharmacology 21:175-185; Carlini et al. (1973)
This paper is a cautionary tale about how science actually works — and how it sometimes doesn't. A positive clinical trial, published by credible investigators with NIH funding, in a peer-reviewed journal, can be buried for decades if the political climate, regulatory framework, and prevailing assumptions are hostile to its implications.
The families who moved to Colorado in 2013 didn't know they were rediscovering a forty-year-old finding. The scientists who ran the modern Epidiolex trials knew — and acknowledged the Brazilian legacy. The vindication was complete, but it came too late for the thousands of epilepsy patients who lived through those lost decades without access to a treatment that had already been shown to work.
“It seemed a very promising approach, but unfortunately, nothing has been done ever since.”
— Raphael Mechoulam
Hebrew University of Jerusalem
Reflecting on the decades of silence after the 1980 trial he co-authored with Carlini
Frequently Asked Questions
Cite this study
Cunha JM, Carlini EA, Pereira AE, Ramos OL, Pimentel C, Gagliardi R, Sanvito WL, Lander N, Mechoulam R. (1980). Chronic administration of cannabidiol to healthy volunteers and epileptic patients. Pharmacology. https://doi.org/10.1159/000137430